124 ALASKAN GLACIER STUDIES
and overridden the interior flat, the new black hornblende moraine might have been* pushed out as far as the old one. One may infer, that the old black hornblende moraine was not the result of a spasmodic advance similar to that of 1906, but either of a much greater advance of this character, or else of a continuous advance maintained for a considerable period of time.
While the most striking feature of the Variegated Glacier in 1909, as compared with its condition in 1905, was the utter change in the moraines, there are other notable differences. One of these was the absence of well-defined ice drainage and of large moulins; but these were developing. Another noteworthy point is the evidence of lowering of ttte glacier surface since the advance of 3:906. This h)as already been discussed in our consideration of the effect of ablation in smoothing the broken glacier surface. It is also-clearly evident by a comparison of the earlier photographs with the 1909 condition. For instance, we reoccupied the sites of 1906 photographs on the south side of the glacier, and from there the lowering of the surface and change from a jagged surface to a rolling one was remarkable. Indeed, we estimate that it has amounted to at least 200 feet,, and probably as much as 300 feet in the part of the glacier lying beyond the mountain valley, though less within this valley; but whether all this lowering is due to ablation is not certain. It is probable that there had been some settling due to further flowage of the glacier after the thickening which the spasmodic advance caused.
As compared with the Atrevida and Haenke Glaciers, while the main changes since 1906 have been the same, Variegated Glacier introduces one point of notable difference. The two other glaciers have reverted to their former completely moraine-veneered condition, the only exception being in the area of clear ice introduced in the Atrevida just outside the mountain front. But Variegated Glacier has not returned to its former condition, and the area of clear ice beyond the mountain front is greatly increased. A partial' answer to the question as to the reason for this difference is undoubtedly that originally there was less thickness of ablation moraine on the Variegated bulb, this being possibly due to the greater stability of the crystalline rocks of its mountain valley as compared with the fissile and crumbling sedimentary rocks which enclose the valleys of the other two glaciers. But this answer is only partial, for the Variegated bulb was almost uniformly moraine-covered in 1905 and now it is not. This moraine has gonesomewhe re. It is not swallowed up in the crevasses, for the bottoms of most of them are visible; and the only conclusion that one can reach is that it has been carried forward. This,, however, raises a new difficulty. Had the red moraine been bodily moved forward a mile, carried on the glacier surface while clear ice from above followed down behind it, the front should have moved forward approximately that amount; but it did not. To account for the increase in the area of clear ice, and the forward movement of the moraine cover without a pronounced advance of the glacier front, seems to us impossible of explanation on any other hypothesis than that clear ice from below has risen up here, and displaced and pushed the old moraine cover forward and to the sides, and that the transfer of ice and moraine has been accomplished by thickening of the bulb rather than by a pronounced forward movement of the front. With reference to the valley part of the glacier this clear ice area had the same relative position as that of the Atrevida. It is possible that in these two cases we have an illustration of the same phenomenon and an explanation of the development of clear ice areas that afterward form interior flats.
The Interior Flat in 1909. In 1909 we reoccupied essentially the same sites as those